To understand how proteins recognize specific sites on double-stranded DNA, and to understand how this binding affects gene function, we have begun an X-ray crystallographic study of the lambda repressor and of the repressor-operator complex. We have already determined the structure of repressor's N-terminal domain (at 3.2 angstrom resolution) and have proposed a preliminary model for the repressor-operator interaction. We propose to refine the structure of the N-terminal fragment of lambda repressor at higher resolution, to continue model-building studies of the repressor-operator interaction, and to study the complex directly by solving the structure of co-crystals. We have already grown crystals that contain the N-terminal fragment of repressor with an 11-base-pair operator fragment and crystals that contain the N-terminal fragment with a full 17-base-pair operator site. Our best co-crystals diffract to approximately 3 angstrom resolution in one direction and to 6 angstrom resolution in the other directions. We will test other crystallization conditions and will use other operator fragments in an attempt to grow better co-crystals. We will also try to grow co-crystals that contain intact lambda repressor and co-crystals that contain lambda cro protein.